Domain informed vision language model for sustainable freight with drayage truck powertrain and cargo classification

Abstract Drayage trucks are a significant emission source in freight-intensive port corridors. Their short-haul and predictable operational patterns make them strong candidates for zero-emission technology adoption. Accurate identification of their powertrain and cargo types is critical for tracking policy implementation, planning infrastructure, and reducing empty-haul emissions. However, conventional methods rely on manually labeled datasets, and few studies address powertrain classification. This study introduces ZeroDray, a zero-shot classification framework that enables a domain-informed vision-language model to identify drayage truck attributes. Domain-informed prompts integrate expert knowledge, visual evidence, and spatial reasoning to generate interpretable predictions with human-readable explanations for transparency and validation. The framework was evaluated on 443 distinct images collected along a highway corridor serving the Ports of Los Angeles and Long Beach and achieved $${F}_{1}$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:msub> <mml:mrow> <mml:mi>F</mml:mi> </mml:mrow> <mml:mrow> <mml:mn>1</mml:mn> </mml:mrow> </mml:msub> </mml:math> scores above 92 percent across 11 powertrain-cargo classes. ZeroDray offers a framework for tracking zero-emission drayage truck adoption and supporting data-driven sustainable freight regulations.